Haoran Lu
Papers
3
Total Citations
13
H-Index
2
About
Haoran Lu is a rising researcher in robotics and embodied AI, whose work focuses on enabling robots to generalize manipulation skills to novel, unseen objects and environments. His core contributions lie at the intersection of few-shot learning, affordance reasoning, and scalable robot learning. In his highly cited work, "Where2Explore," Lu pioneered a few-shot affordance learning framework that allows robots to infer how to manipulate articulated objects—like cabinets or drawers—even when they have never encountered that object category before, addressing a fundamental bottleneck in robotic generalization. His subsequent work on "RoboVerse" provides a unified platform and benchmark for scalable robot learning, aiming to bridge the gap between simulation and real-world deployment. Additionally, "ImageManip" introduces affordance-guided next-view selection, enabling robots to actively choose camera angles to improve manipulation success. With over 13 citations across his early-career publications, Lu’s research is already shaping how robots can learn from limited data and interact with diverse, unstructured environments—a critical step toward capable home-assistant robots.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3